Thanks Francesco,

This was very helpful. As far as I understand, in order to prevent noise, you 
have encouraged subjects not to blink excessively. Some subjects might have 
taken this very seriously and blinked only few times per minute. Other subjects 
might have only mildly suppressed their blinks in order to provide good signal. 
Do you think this might have happened? Do you think the blinks are not 
spontaneous, but semi-inhibited?

This does not represent any issue for resting state analysis (apart from slight 
effect of blink suppression on some cortical regions as documented with fmri). 
But for my research purposes, I would prefer natural behavior. It would be very 
important for me to understand this before I spend time in analysis.

Is there a readme file describing what those MEG pipeline scripts aim to do? In 
the script headers, purposes are not defined clearly. I can try to estimate 
what they do from the names. When I dig in I can estimate a bit more about they 
do. It would be nice to have a simple file summarizing their purposes.

I also wonder what .vs extension stands for in the ica output txt files.
Does it mean  'visually screened/checked'?

Could you also emphasize the standard MEG processing procedure? I am a bit 
confused about the 2s epoching in fieldtrip. Is it the case that after 
automatic and manual selection of the bad segments and channels, those trials 
coinciding with the bad segments are replaced by NaN and bad channels are taken 
out? When I look at the IC output figures there are spaces between some data 
points. Could you clarify how chopping the bad segments and ICA work out in the 
pipelines?

Thanks!
Baris



From: Francesco Di Pompeo [mailto:[email protected]]
Sent: Monday, March 14, 2016 11:58 AM
To: Demiral, Sukru (NIH/NIAAA) [F] <[email protected]>
Cc: [email protected]
Subject: Re: [HCP-Users] Resting State MEG instructions and reports

Hi Baris,

On Mar 11, 2016, at 11:13 PM, Demiral, Sukru (NIH/NIAAA) [F] wrote:


Hi there,

I was wondering what type of instruction was given to MEG subjects for the eyes 
open resting state recordings. Was it documented?

Yes you can find all the details in the HCP reference manual 
(HCP_S900_Release_Reference_Manual.pdf page 50). As you can find there "during 
the scan the subject is supine, padded for comfort, instructed to relax with 
eyes open and maintain fixation on a projected red crosshair on a dark 
background (presented in a darkened room)"
You can find a very detailed explanation of all the acquisition steps and 
instructions in the Appendix IV HCP_S500+MEG2_Release_Appendix_IV.pdf.


Did anyone in the MEG group do a quality assurance of whether subjects really 
kept their eyes open, and reported that?

Yes, we had a set of quality checks in order to address this issue. First of 
all the acquisition team was monitoring the subject behavior during 
acquisition. Moreover, at the end of the icaclass pipeline a visual inspection 
of the ICA results and of the ECG and EOG signals has been done for several 
purposes. During this step, the EOG signals were monitored as well as IC 
corresponding to eyes movements.

I looked at some subjects, and realized that they did not blink much; some ways 
they might have kept their eyes closed, making the resting state 'eyes closed' 
not eyes open.

I have done the QC of all the subjects and I can say that there are subjects 
with poor eyes movements and it is normal. According to my experience, some 
subject, probably misunderstanding the instruction, can try to stay with eyes 
open during all the acquisition and do not blink if not really necessary. If 
you have some particular case please let me know the subject number.

In addition, when I opened the raw file (c, rfDC) with SPM, it shows the EOG 
channel labels differently than when I open the file with sat Brainstorm. I 
wonder what the best way is to visualize and export the EOG vertical and 
horizontal channels (and also ECG) time series in fileldtrip. Or to put it 
another way, how can I get the finalized physiological channels'  time series 
(with their correct labels) ? I found the script "hcp_exgmontage.m", shall I 
use this one?

You can use part of the icaclass pipeline if you would like to use 
matlab/fieldtrip.
Around line 150 you can find instructions like

    montage       = hcp_exgmontage(subjectid, experimentid, scanid);
    dataELEC=ft_selectdata(dataraw,'channel',montage.labelorg);
    dataELECnew=ft_apply_montage(dataELEC,montage);

If you would like to process the ref data has done in the HCP release (i.e 
filtering, bad channels and bad segments) you can use all the code you have in 
icaclass pipeline until line 232 by avoiding the part of the ICA decomposition 
which is time demanding.
The resulting ref channels are in the variable ref_data (line 232). Of course 
you can change the parameter has you want.

Also, Is there an ICA component written out which is classified as blink or ECG 
component?

Yes. The result of the icaclass pipeline contains all the components and the 
classification of them in brain, generic artifacts and eog/ecg related ones.
For each session you can find the eog and ecg related artifacts in the 
icaclass_vs output files (177746_MEG_3-Restin_icaclass_vs.txt as an example). 
You can find the same info in the icaclass_vs.mat file in the class variable.

This is the content of the ASCII files

vs.good = [17 32];
vs.bad = [];
vs.total_ic_number = 37;
vs.brain_ic_number = 30;
vs.brain_ic = [2 3 4 5 6 7 8 9 10 11 12 13 14 15 18 19 20 21 22 23 25 27 28 30 
31 33 34 35 36 37];
vs.ecg_eog_ic = [1 16 26 29];
vs.flag = 1;
vs.physio = [];
vs.brain_ic_vs_number = 32;
vs.brain_ic_vs = [2 3 4 5 6 7 8 9 10 11 12 13 14 15 17 18 19 20 21 22 23 25 27 
28 30 31 32 33 34 35 36 37];

The variable you are interested in is vs.ecg_eog_ic which contains the indices 
of the independent components related to ecg and eog artifacts.

Let me know if you have any further comments or questions,

Best,
Francesco





Thanks very much in advance,

Baris


Sukru Baris Demiral, Ph.D.
Laboratory of Neuroimaging (LNI)
National Institute on Alcohol Abuse and Alcoholism (NIAAA)
National Institutes of Health
10 Center Drive, Room B2L124
Bethesda, MD 20892-1013
Office: 301-402-5861
Cell:    314-681-0655


_______________________________________________
HCP-Users mailing list
[email protected]<mailto:[email protected]>
http://lists.humanconnectome.org/mailman/listinfo/hcp-users

Francesco Di Pompeo, PhD
Institute of Advanced Biomedical Technologies and
Department of Neuroscience and Imaging
University of Chieti "G. d'Annunzio"
Via dei Vestini - Campus Universitario
66013 Chieti - ITALY
ph: +39-0871-3556907
fax:+39-0871-3556930



_______________________________________________
HCP-Users mailing list
[email protected]
http://lists.humanconnectome.org/mailman/listinfo/hcp-users

Reply via email to